After a significant rally, companies providing infrastructure for the AI boom—often referred to as "pick-and-shovel plays"—have recently faced a sharp market correction. However, analysts argue that this does not undermine their strong year-to-date performance, which has been supported by robust fundamentals.

According to trading desk reports, a recent UBS note highlights that while valuations for these industrial stocks have expanded, market expectations for their long-term growth remain relatively modest compared to the lofty projections for tech giants.

In a report dated November 14, UBS tracked a portfolio of around 60 global AI-related industrial stocks, which saw an average decline of about 5% in the previous trading session, with some individual stocks dropping as much as 10-20%.

Despite the pullback, the group’s strong year-to-date performance remains intact. On a market-cap-weighted basis, the portfolio has delivered a 41% return this year, significantly outperforming broader markets and nearly doubling the performance of the "Magnificent Seven + Broadcom" tech leaders. This rally has been underpinned by tangible improvements in business performance, including soaring cash flow return on investment (CFROI) and accelerating asset growth.

The robust stock performance has driven valuation expansion. UBS data shows the portfolio’s economic P/E ratio has risen from around 25x a year ago to nearly 35x today. While this represents a significant premium over broader markets, it remains below the valuations of leading tech giants, reflecting optimism about future earnings growth for these AI infrastructure providers.

**Strong Fundamentals Support Performance** The outperformance of AI infrastructure stocks is not speculative but backed by real demand and earnings growth. The UBS report emphasizes that the AI boom remains the dominant long-term theme driving industrial sector performance, with capital expenditures from big tech firms serving as the primary catalyst.

Key data points cited in the report include:

- Meta has raised its 2025 capex forecast from $66-72 billion to $70-72 billion, signaling that 2026 capex growth will be "meaningfully larger" than 2025. - Alphabet increased its 2025 capex estimate from $85 billion to $91-93 billion, with "significant growth" expected in 2026. - Microsoft noted that due to accelerating demand, its FY2026 capex growth will outpace FY2025. These three companies alone generated $60.8 billion in free cash flow in just one quarter.

Amit Mehrota, head of UBS’s industrial sector research, stated there are no signs of a slowdown in AI-related capex, which bodes well for all industrial firms tied to the AI theme.

Demand is also broadening beyond data center hardware like racks, cables, and cooling solutions. The massive power requirements of AI are emerging as a new growth driver. The International Energy Agency (IEA) predicts global data center power consumption will double by 2030, creating opportunities for companies ranging from Vertiv (VRT), a leader in data center equipment, to traditional industrials like Caterpillar (CAT), which benefits from turbine demand.

Corporate data indicates a turning point in operational performance for these AI infrastructure stocks. The portfolio’s CFROI has surged from mid-to-high single digits in the early 2000s to over 10% recently. Meanwhile, economic profit has grown at a compound annual rate exceeding 50% over the past four years. More than half of the companies rank highly in CFROI improvement, with most also scoring well on "growth factors" measuring future cash flow potential.

**Valuation Expansion vs. Market Expectations** Alongside strong stock performance and fundamental improvements, valuation multiples for these AI enablers have indeed expanded. UBS’s HOLT model shows the portfolio’s economic P/E ratio has risen from about 25x in November 2023 to nearly 30x by November 2024, now approaching 35x.

While still below the "Magnificent Seven + Broadcom" group, this valuation is significantly higher than broader market benchmarks, reflecting investor optimism about future growth. But how much growth is already priced in?

Using its HOLT framework, UBS reverse-engineered market expectations and found that long-term growth projections for these industrial stocks are less aggressive than their price action suggests. The analysis indicates the market is pricing in an average long-term (4-10 years) sales CAGR of just 6% for these ~60 industrial stocks.

In contrast, the same framework applied to the "Magnificent Seven + AVGO" group shows a market-implied long-term sales CAGR of 9%. This suggests investors hold more conservative growth expectations for AI infrastructure providers compared to pure-play tech firms.

**Stock Divergence: High vs. Low Expectations** Despite the modest average outlook, market expectations vary widely across individual stocks, creating clear bifurcation. The UBS report highlights three case studies: Bloom Energy (BE), First Solar (FSLR), and Schneider Electric (SCHN).

- **Bloom Energy (BE): High-Expectation Play** Due to its solid oxide fuel cells that can rapidly provide on-site power for grid-constrained data centers, BE’s stock has surged over 400% this year. The market prices in a long-term sales CAGR of ~14%, the highest in the portfolio, indicating lofty investor hopes—and execution risks.

- **First Solar (FSLR): Low-Expectation Extreme** Despite benefiting from clean energy demand and U.S. policy support, FSLR’s implied long-term sales CAGR is near zero (<1%), reflecting concerns over historical volatility and potential margin pressure from changing tax credit policies. This ultra-low expectation may present opportunities for contrarian investors.

- **Schneider Electric (SCHN): Moderate Middle Ground** As a leading European supplier of AI data center infrastructure and power solutions (20% of revenue), SCHN has a strong track record of value creation. The market prices in a ~5% long-term sales CAGR, placing it in the portfolio’s "middle tier"—a neutral stance given its stable market position and strategic partnerships (e.g., with Nvidia and Microsoft).